Wall Street’s debate over Bank of America’s purchase of Countrywide Financial is not whether it’s a good or bad deal—but whether it will be C.E.O. Ken Lewis’ last deal. “If this doesn’t work, at some point shareholders are going to demand Lewis’ head,” says Josh Rosner, managing director of Graham Fischer & Co., a New York research consultancy that specializes in mortgage finance.

But in keeping with the market axiom that any viewpoint unanimously held is wrong by definition, writing off Lewis so fast may be a mistake. In fact, there’s a decent—and obviously contrarian—case to be made that the Countrywide deal plays to Lewis’ strengths, if he can survive long enough to show them off.

First, Lewis is an old hand at big deals. Bank of America is the product of more than 3,000 mergers, and Lewis has worked on many of them, including the epic 1998 union of NationsBank (known as North Carolina National Bank when he started there in 1969) and Bank of America, California’s largest bank. It’s telling that most analysts were initially skeptical of the two Lewis megadeals that clinched Bank of America’s place as the country’s first (and still only) truly national consumer bank: the $48 billion acquisition of FleetBoston Financial in 2003 and the $34 billion purchase of credit-card specialist MBNA in 2005. In both cases, B of A was roundly criticized for overpaying. In the end, Lewis and his team squeezed so much value out of FleetBoston and MBNA in integrating them with B of A’s existing operations that investor opinion on the deals swung strongly to the positive.

Lewis also showed unusual resolve in sitting on his checkbook even as rivals were paying top dollar to buy mortgage lenders by the dozens in the 1990s. He was deeply ambivalent about the mortgage market, telling colleagues that he “loved the product and hated the business.” Although he acknowledged that the home loan was a “cornerstone of the customer relationship,” he considered mortgage lending unnecessarily risky. Not only did housing markets go from boom to bust all too often, but the structure of the industry was too risky because banks outsourced much of their lending to brokers or others.

As soon as Lewis moved up to C.E.O., he began restructuring Bank of America’s mortgage-lending operation with the dual aim of reducing risk and forging a closer, more lucrative relationship with his customers. Within a year, the bank had stopped making subprime loans and buying mortgages from other lenders. It also found ways to attract better-quality buyers: It trained about 10,000 personal bankers in its branches to sell home loans directly to consumers, along with checking accounts, credit cards, and certificates of deposit. This might not sound like much, but other banks relied on referrals from homebuilders and real estate brokers, which meant that they had less control over customer quality.

Lewis’ conservatism was tested during the housing boom that lasted from 2003 to 2006. Countrywide and other lenders fed the frenzy by amending conventional standards of creditworthiness to accept almost anybody willing to fill out a loan application (ability to repay optional). As lending volume exploded, the issue of whether B of A should jump back into the subprime game sparked fierce internal debate. Customers by the millions were going elsewhere for mortgages and taking their banking relationships with them. “You don’t want to be in a position where you have to say no to customers,” says a B of A executive who found himself on the other side of the issue from his C.E.O.

Lewis’ intransigence became prescience when the credit-market convulsions of mid-2007 brought the subprime house of cards tumbling down on its architects, including Angelo Mozilo, Countrywide’s 69-year-old co-founder and C.E.O.  While mortgage originations across the industry for the year dropped 15 percent, Bank of America’s volume jumped 21 percent, to $93 billion. More important, the $275 billion worth of mortgages the bank held on its books remained solidly profitable even as soaring rates of default and foreclosure pushed many subprime specialists into bankruptcy. For 2007, B of A’s net mortgage charge-offs totaled $57 million, a mere 0.02 percent of its portfolio.

For Lewis to acquire the company that underwrote so much of the subprime sludge now polluting the global financial system isn’t quite the betrayal of principle it might seem. Until the late-career meltdown of Mozilo’s credit judgment, Countrywide was an enterprising, well-managed company that for three decades had subsisted on exactly the sort of conventional home loans to which Lewis limited B of A. Countrywide hasn’t made a material subprime loan in nearly a year, and Mozilo is a self-correcting problem—for Lewis, anyway. The lavishly paid, superannuated poster boy of subprime excess will retire as soon as the sale closes, leaving a horrendous legal mess—including his Friends of Angelo V.I.P. program—that could take B of A a few years and a whole lot of money to clean up.

What investors fear above all is that B of A’s $3 billion purchase price will turn out to be the mergers-and-acquisitions equivalent of a modest down payment on a grand old mansion that’s now so riddled with termites and toxic mold that the bank will have to spend an additional $10 billion to $30 billion to bring it up to code. The money pit consists mainly of the mortgages that Countrywide carried on its books as an investment at the end of the first quarter. As of March 31, 4.16 percent of those mortgages were in default or foreclosure, compared with 1.65 percent six months before and just 0.66 percent at the end of 2006.

This makes for a trend line that big investors might well try to fashion into a noose for Lewis if housing markets like California and Florida fail to stabilize soon. In closing on its purchase of Countrywide, B of A must mark to market its $95 billion mortgage portfolio to something approaching a real price—a process that analysts expect to be ugly. “A year and a half from now, Ken Lewis is going to sit down and tell somebody over drinks, ‘I wish I had walked away from Countrywide,’ ” says Paul Miller, an analyst at FBR Capital Markets, who predicts a $28 billion markdown, give or take a few billion.

People inside B of A, though, insist that Lewis is holding firm. In addition to being what is still the busiest mortgage lender in the country—Bank of America is fifth—Countrywide also remains the country’s largest mortgage servicer, handling the billing and other administrative tasks for nearly $1.5 trillion in loans. Servicing is a business of scale and data-processing precision in which Countrywide’s growth has long set the pace and Bank of America has lagged. Shifting B of A’s $530 billion portfolio onto Countrywide’s servicing platform could result in significant operating economies and better profit margins.

Like all gamblers, Lewis needs to be lucky as well as good. Bank of America’s C.E.O. excels at the tricky, painstaking work of postmerger integration but exerts no control whatsoever over the great X factor on which his job security apparently now hangs: housing-market fundamentals. He has told analysts that all is going according to plan. “All I can say is nothing has happened that is out of the boundaries of what we contemplated when we did the deal,” he said during a recent Deutsche Bank conference call.

On the call, Allen Puwalski of Paulson & Co. pressed Lewis, saying that housing prices could well decline much more than the consensus forecast of 25 percent.

“If that’s the case,” Lewis retorted, “we’ll be worried about Countrywide, but we’ll be worried about a lot of other things too—and not just at Bank of America.”